From a single team experimenting on millions of people using it daily, the Scrum retrospective is rich in lessons. On this occasion – and it is rare enough to be highlighted – the co-creator of Scrum, Jeff Sutherland, spoke. He trained me (already 17 years ago!!), and I always align with his vision. He’s pretty old now, so I don’t miss an opportunity to listen to him. If you did not have the opportunity to participate in his conference, here are the elements I took away from it.
Review Of 30 Years Of Scrum
30-year retrospective: Scrum was born out of empiricism… and invites us to be in this state of mind every day. In 1993, in Massachusetts, a team in a computer software company started an experiment with:
- One sprint: 1 month at a time. The duration will be refined later because they will realize that it is too long;
- 1 team: it will be completed later by adding skills via the recruitment of engineers from the competition;
- One objective: create new generation software to maintain the company’s leadership in the face of the meteoric rise in competition in the context of the explosion of the IT sector at the time.
What struck me about this experience was that the team, despite being under tremendous pressure, still took the time to work on productivity:
- 50% of working time was devoted to creating the product,
- and 50% to productivity.
So every morning, the team read a study or an article (for example, on the Toyota system or an academic essay, etc.), tested it during the day, and did a retrospective in the evening. One day, they read an article about Borland’s team, at the time, the best development team in the world and by far the fastest. They asked themselves, “What is this group doing that we aren’t doing?” Reply: “a day to day gathering!” The Borland group met daily to see what to do today to “move the framework” and convey quicker. They, consequently, continued this training, which is today a necessary piece of Scrum.
At the end of 1993, the work organization of this first team was named “Scrum.” This framework has since developed in all sectors of activity, from industry to finance. My opinion on this feedback: the genesis of Scrum, which was enriched little by little by experimenting and keeping what works and removing everything that is an obstacle to productivity, is typically agile!
This invites us to refocus on the “Secret Sauce” of Scrum because this brings results: what do I need to change in the system to deliver faster? This reflection, in which the whole team looks at how it works and how it can improve it, is essential. We do not hold a daily meeting to report on tasks but to identify the most relevant action to carry out during the day to achieve the objective. And for the record, 30 years later, the product – created by this first pioneering Scrum team – is still in use. Better than that, he is still considered one of the best today!
And For The Future? Chat GPT Replaces My Scrum Master?
Here again, to sketch the future and glimpse the place of AI in teamwork, Jeff Sutherland (and JJ Sutherland, his son taking up the torch) suggests that we do… Scrum! Just as back then they watched the best teams in the world, today they watched how the best chess players integrated machines into competitions that allow human/computer games. It emerges that the big winners are neither humans alone against machines nor machines alone against humans, but human teams who integrate devices as team members.
Based on this observation, with several groups, Jeff Sutherland is today experimenting with having Chat GPT as a Scrum team member. Starting from the observation that there are things that humans do better than machines and things that machines do better than humans, the thinking is the same as that of the first Scrum team who asked themselves, “What does Borland do? better than us?”.
Updated, this reflection becomes “What do machines do better than us, and how could that help us?”. In their test team, the answer to this question is that machines are better than us at analyzing millions of lines of computer code. So they put several years of computer code into ChatGPT, which studied it and said the best thing to do immediately. In this example, the AI helps the team to be even more cross-functional by adding a skill (here, the analysis of millions of lines of code).
If you are familiar with software, you know “Machine Learning,” which is how a machine becomes efficient. For example, Google, to optimize its advertisements, will have an initial period of machine learning: it launches the ad on a pre-defined audience with several texts and images, then adapts it according to the clicks and statistics collected to refocus on the audience combo + text + image which works best.
Machine learning is an iterative process that works very well with Scrum itself. The interactions of the Scrum team with the machine will influence the latter to push it towards agile values and thus make the devices “human-centric” and thus reduce the risk associated with AI. The purpose of this article is not to enter into the debate on the risk linked to AI, and there are people better placed than me to talk about it. But it seemed interesting to me to tell you about this current test. This could become our daily reality, so let’s be prepared.
Read Also: YouTube Video Campaign, Five Mistakes That Undermine Your ROI
